My name is Nils Everling. I live in Stockholm, Sweden where I work as a quant portfolio manager within the systematic equities group at AP4. I do a bit of hiking and I have a passion for film (see my Cinetrii project or follow me on Letterboxd).
About Me
Education
I studied at KTH and received my M.Sc.Eng in computer science in 2017. For my master thesis I devised a way of estimating sector risk of equities using topic modeling on earnings call transcripts.
Projects
Cinetrii.com caters to my own interest in film. It maps out the lineage of films - which previous works have inspired a given film, and what that film has in turn inspired. It does so by aggregating references and name drops by film critics, who are generally knowledgeable in film history. I built Cinetrii in 2016 and sporadically work on it.
Huffington Post called it "the best movie recommendation webapp we’ve come across".
Featured In:
Cinetrii started out as a project in Java but has since been migrated to Python, which is far more suitable for NLP 🙂
Tech I like
- Streamlit - I have been using it since I saw it on Hacker News in 2019. Streamlit makes it easy to develop and deploy webapp tools and monitors for internal use. Before, python data scientists who wanted to bring a notebook to production had to use Flask, and I'd rather not get bogged down with frontend code.
- Newspaper4k - Continuation of Newspaper3k. Article text retrieval from news websites. Super useful for Cinetrii.
- Zipline-reloaded - Actively maintained equity backtesting engine originally built by Quantopian before they merged with Robinhood. Perhaps not the first choice for a backtester in 2024, but coupled with an institutional data feed the Pipeline API is quite powerful.
- AI? Copilot helps me with sqlalchemy connections, Dockerfiles and all the stuff I don't touch on a regular basis. I built this personal web site using a lot of help from Claude.ai.